Inspiration

Simulation is important in industrial development. We plan to do the optimization of a simulated test-system to reduce measuring time and increase product output.

What it does

The optimization code maximizes the output per 24h

How we built it

We wrote a Python testing script based on the DUT simulator source code.

Challenges we ran into

  • We spent the first night to understand the source code of the simulator.
  • The generated samples from the DUT simulator is random. So it's hard to find the pattern/correlation between measurements.
  • Lack of theoretical knowledge in data science and machine learning
  • Nobody really knows Git

Accomplishments that we're proud of

  • Group work in a team of all Hackathon beginners.
  • The measurement time is reduced by (a few) percents!!!
  • No black-box ML/DL model used!

What we learned

  • Git SCM via GitHub
  • Python framework: pandas, numpy, scipy
  • Pycharm IDE

What's next for DUT Me

  • clean the messy coding style of Python test script
  • make an accurate DUT simulator based on the real device
  • Create a database of simulated DUT with different number of measurements (N) and number of measurement ports (P)
Share this project:

Updates